Society, Technology, and the Sustainable Provision of Electricity

(1)

Society, Technology, and the

Sustainable Provision of Electricity

A sociotechnical perspective on developing energy infrastructure

Seth Townley

Master thesis in Culture, Environment and Sustainability

Centre for Development and Environment UNIVERSITY OF OSLO

June 2016

(2)

II

Society, Technology, and the Sustainable Provision of Electricity: A sociotechnical perspective on developing energy infrastructure

http://www.duo.uio.no/

Print: Reprosentralen, University of Oslo

(3)

III

Abstract

A direction which sustainable development in energy systems could take is towards electricity ‘co-provision.’ Here, centralised fossil-powered generation is partly replaced by decentralised, renewable micro generation infrastructure supplementing a leaner, more dynamic national grid fuelled by diverse range of sustainable power sources.

Pursuing this path would require a major overhaul in existing technical systems of energy provision. The common theme which runs through much of the diverse literature on technology change emphasises the sociotechnical nature of change processes, which is to say that it treats technology and technical systems as the material embodiment of social choices. This means that the process of developing energy infrastructure into a sustainable system of electricity co-provision cannot simply be ‘enabled’ by technology, conceived of as an exogenous variable ‘impacting’ society. It is, in fact, a process of endogenous societal change: a sociotechnical transition. This thesis examines the dynamics of this transition with regards to electricity co-provision. In doing so, it presents and seeks to problematize simplistic discourses on technology change, and to draw back from the narrow conception of technical innovation found in these discourses. It analyses existing systems of energy provision alongside individual energy project case studies, which represent novel sociotechnical configurations of artefacts, institutions, and actors in electricity co-provision experiments. It applies sociotechnical transitions theory and its multi-level perspective in order to link the micro and the macro levels within this analysis. The thesis attempts to highlight both the promise in the emerging configuration of co-provision technologies and actors, but also the doubt that hangs over any vision of a sustainable sociotechnical transition. To do this it draws attention to the multiple paths along which sociotechnical change could unfold, but which are ‘closed off’ by a number of predominantly political factors. It is shown that, ultimately, the future shape of technical systems is not simply a question of technology but of social choice. Trying to better understand who is able to influence these choices, and who is locked out of the process, is an important area of enquiry in environment and development studies. This thesis aims to contribute to this area of enquiry, specifically as it relates to the sustainable provision of electricity.

Key words: electricity co-provision, sociotechnical transition, sociotechnical configurations, innovation trajectories/ pathways, sustainable development

(4)

IV

(5)

V

Acknowledgements

I would like to extend sincere thanks my supervisor, Desmond McNeill, whose patient guidance has been invaluable as this thesis, falteringly at first, has taken shape. Also, to my partner, Katy, for graciously putting up with my now year-long obsession with the esoteric minutiae of sociotechnical energy systems, and for always supporting everything I do. Thanks also to Gudrun Helland, our fantastically supportive study coordinator at SUM, for cheering me and everyone else along and always being on hand to answer questions; Tanja Winther for taking time to discuss my early project proposal with me; Andy Edwards of Bioregional for talking to me before my idea was even an idea, but helping me to make contacts necessary for my fieldwork all the same; and to all of my informants for taking the time to share their knowledge and expertise. And last, but most certainly not least, I want to thank all of my fellow SUM master’s students who have shared this experience with me. Completing this thesis has been somewhat of an emotional rollercoaster, but I have only very happy memories of all the lunches and coffee breaks we have taken together along the way.

(6)

VI

List of figures

Figure 1: UK Electricity Generation Energy Mix 2010-14. Source: DECC 2015a ... 29 Figure 2: The UK electricity supply chain with annotations detailing the institutional arrangements for the supply of electricity in the UK. Design: the author, based on information from Harris 2006; Elexon 2013 ... 32 Figure 3: An example of the daily load profile in the UK. Reproduction. Source:

www.greycellsenergy.com ... 34 Figure 4: Total energy consumption (includes home-installed gas boilers) in the UK by user type. Reproduction. Source: DECC 2015c ... 37 Figure 5: Actuals and estimates predicting the average trend in the breakdown of a UK electricity bill over time. Reproduction. Source: Ofgem ... 48 Figure 6: National trends in wind generation as a proportion of total consumption across the UK, Germany and Denmark. Sources: DECC 2015a; BEW 2016; Energistyrelsen 2016 ... 54 Figure 7: Organisational map of actors in empirical study. Design: the author. Source:

field study and respective organisations’/ projects’ websites. ... 76

(9)

IX

List of tables

Table 1: Energy storage technologies for integrating into centralised and distributed electricity generation and generation-transmission systems. Source: Renewable Energy Association 2015. ... 41 Table 2: An exemplar sociotechnical configuration of electricity co-provision artefacts, social relations, and the institutional structures in which these patterns are embedded.

Design: the author. Source: field study (but see also Watson 2004, Table 1, 1983). ... 69

(10)

(11)

1

1 Introduction

1.1.1 Effecting a sustainable energy transformation: a question with a technological answer?

One day soon there will be no need for fossil fuelled power stations, for nationwide electricity grids, for conventional energy companies. ‘Smart’ technology will allow consumers of energy to produce their own electricity and heat with renewable micro generation and energy storage managed at a small, localised level. All that remains to be done is to establish the correct market conditions for investor confidence to reach critical mass, becoming the engine of innovation that will drive the transformation towards sustainability. Obsolete hydrocarbon conglomerates will be superseded by the more dynamic and forward-looking telecommunications, or ‘telco,’ industry, with its proven track record in delivering rapid, transformative change with ‘game changing’

technology. People will no longer be connected to communications networks merely by their personal computers and mobile handsets: all energy-using appliances, automobiles, and buildings will possess the property of ‘connectivity,’ autonomously sending and receiving data as a means to optimise performance on behalf of their human masters’

welfare. Following this technology-enabled revolution, services such as energy will be just another ‘app’ in the ‘Internet of Things.’

The above is now a familiar utopian narrative. It was repeated, once again, in a series of claims made at a recent ‘breakfast dialogue’ on sustainable transformation held in Oslo.¹ The fact that the claims were made, in this instance, by a spokesperson for a large telco may make the particulars less convincing; the academics with whom this spokesperson was in dialogue at breakfast time appeared sceptical, to say the least. This fact makes the vision no less compelling, however: aside from the real and imminent danger posed to societies by anthropogenic climate change, associated with emissions from fossil fuel combustion (IPCC 2014), the centralised model of commoditised energy provision prevalent in almost every country is currently failing to grant access to electricity for 18% of the world’s population (IEA 2015). Even within countries such as the United Kingdom, with ostensibly universal access to energy, high cost to income

1 Paraphrased here from a recording available to watch here (see in particular 00:39:00 and 01:08:00 onwards): https://www.youtube.com/watch?v=0HPFaK7Kubg&feature=youtu.be last accessed 25.04.16

(12)

2

ratios exclude millions from affordable access to sufficient heat and electricity in homes, compounding existing public health and social problems related to inequality in these countries (Liddel and Morris 2010). Transforming systems of energy provision along the lines supplied by the ‘technology-enabled revolution’ narrative would not only mitigate climate change and environmental degradation, it could improve the material and social circumstances of billions of people globally. What then, is missing from this narrative? Why has it not yet reassured the environment and development research community that the question of how to deliver sustainable energy for all is well on its way to a technological answer?

It is true that the narrative—when presented by the telco industry in the terms used here—lacks the level of objectivity and critical self-reflection that might lend it more credence in sustainable development research discourse. It does not, however, present a technological vision that is far beyond the realms of what is widely acknowledged to be already technically possible. A combination of small scale renewable generation technologies, distributed electricity storage, and information technology (IT) make it possible to form electricity generation-transmission networks on a scale of a single household, neighbourhood community, or municipal area (Lasseter et al 2002; Markvart 2006; Zamora and Srivastava 2010). In the context of international sustainable development, the use of these so-called microgrids, particularly in rural areas, has already been shown to be a promising advance in improving access to clean energy, poverty reduction, and also gender equality (Ulsrud et al 2014; Ulsrud et al 2011;

Standal 2015). In countries such as the UK, ‘community energy’ projects that use this technology can already be observed lowering consumption through efficiency savings and changes in user practice, increasing renewable generation, and improving welfare by increasing the level of participatory democracy in energy provision (Hielscher et al 2011; Kunz and Becker 2014).

What is most unconvincing about this ‘technology-enabled revolution’ narrative, then—

aside from the hint of commercial sponsorship by the telco industry—is not the technological vision it presents of the future, the ends of a transformation in systems of energy provision: it is, rather, its account of the means by which this transformation will be effected. The implicit assumption of the ‘technology-enabled revolution’ narrative is that the existence, ‘out there,’ of the required technology is sufficient to, with investor

(13)

3

backing, ‘impact upon’ society in such a way as to deliver the vision that the narrative describes. The market conditions created by governments through subsidies, regulation, and so on will affect the speed and economic efficiency with which this impact takes effect, but the direction and inevitability of change were basically assured by the invention of ‘smart’ technologies.² What this account of society-transforming technological change fails entirely to incorporate is an understanding of the inherently social nature of technology and technical systems.

The academic literature on technology change offers varied accounts on diverse aspects of this topic, but there is a common theme that runs through much of it, which emphasises the sociotechnical nature of change processes. Culture, politics, and economics all inform decisions that lead to the design, development and diffusion of new technologies, the organisation of artefacts, institutions and infrastructure, and the way in which society perceives of, links and uses these components of utility service- providing systems. In short, technology and technical systems are the material embodiment of these social choices. These choices, in turn, are configured by the socio- material environment created by the design of existing artefacts, the services technology and technical systems already provide, and the specific manner in which these services are produced and distributed (Hughes 1983; Nye 1990; Rip and Kemp 1998; Shove et al 1998; Coutard 1999; Veerbeck 2006). The assumption about the nature of technology change which is implicit to the ‘technology-enabled revolution’ narrative is, then, characteristic of a common flaw in approaches to this subject. In the words of science and technology scholar, Bruno Latour:

“We know how to describe human relations, we know how to describe mechanisms, we often try to alternate between context and content to talk about the influence of technology on society or vice versa, but we are not yet expert at weaving together the two resources into an integrated whole”

(Latour 1991, 111).

This thesis, then, examines further the sustainability transformation in energy provision that is envisioned by the ‘technology-enabled revolution’ narrative, but with a treatment that acknowledges the inherently social nature of technology and technical systems. It

2 ‘Smart’ as an adjective for technology is vernacular usage which does not follow a strict technical definition. In relation to “devices or machines” it can mean “appearing to have a degree of intelligence;

able to react or respond to differing requirements, varying situations, or past events; programmed so as to be capable of some independent action; (in later use) spec. containing a microprocessor” (OED 2016).

(14)

4

understands the system of energy provision described by this narrative to be the product of a process of sociotechnical change; a gradual transition away from the centralised, bulk supply arrangements in use today towards a renewed system of provision with unrecognisable technological, infrastructural, and institutional arrangements. Far from being a question simply of markets and technology investment, this is a societal transition (Seyfang and Smith 2007). The system of technological artefacts currently relied upon for the provision of energy are necessarily replaced during this transition, but also transformed over time are the social relations governing the production, distribution, and use of energy, cultural expectations and meanings attached to attendant technologies, the practices of energy consumers, and sociopolitical commitments to certain policy and technological innovation trajectories.

The aim of this thesis is not to forecast, however. It does not seek to describe a neat series of hypothetical social changes that will lead straightforwardly to an imagined sociotechnical future. It attempts, rather, to do almost the opposite: to problematize simplistic discourses on technology change, including the discourse in which the

‘technology-enabled energy revolution’ narrative is embedded; and to draw back from the narrow conception of sociotechnical innovation as an “implacable force moving through history” (Nye 1990, ix) along a set trajectory that is inevitable and ‘pre- ordained’ (Stirling 2009). Through an analysis of existing sociotechnical systems of energy provision, and the application of sociotechnical transitions theory derived from historical processes of sociotechnical change, this thesis is an attempt to highlight the doubt that hangs over the sustainability transformation described by the ‘technology- enabled revolution’ narrative; and, indeed, that hangs over any vision of sustainable transformation. It seeks as well to draw attention to the multiple paths along which sociotechnical change could unfold, but which are occluded by simplistic discourses on technology change, and to make clear the political nature of these dynamics. The purpose of attempting to do all of this is to show that a better understanding of the social and political forces that shape technical systems is essential to environment and development research if any vision of a sustainable transformation in these systems—no matter the specifics of any such vision—is ever to be realised. At stake is not simply the vital utilitarian requirement to reduce carbon emissions, but also questions of social and economic justice: questions about who is able to participate in making the decisions that shape the sociotechnical assemblages which constitute our common future.

(15)

5

1.1.2 Thesis outline and research questions

The empirical analysis in this thesis, then, is—using examples from the UK—of technology in use. That is to say, it examines a specific category of technology, namely the electricity generation, storage, and information ‘control’ technologies that feature in a vision of a decentralised system of sustainable energy provision. However, it defines this technology by its social application, as opposed to purely in socially abstract engineering terms. Chapter 2, A Balancing Act, begins this analysis by doing a number of things. One of these is to define this technology category in broad strokes. To do this it shows how a part of the transition towards a decentralised and sustainable system of energy provision is a change in the relationship between supply and demand. This can be achieved through demand becoming more dynamic and reactive to variation in supply, and also—or alternatively—by consumers of electricity producing some of their own supply. This alternative model of electricity provision—in which some or all of supply is, effectively, moved closer to demand—is well defined in the literature as a model of co-provision (Watson 2004). Chapter 2 defines the technological category under analysis in these terms, as technologies which mediate a closer relationship between supply and demand: co-provision technologies.

In showing how the move towards electricity co-provision fits into a sustainability transition, Chapter 2 presents some quite dense fine grain detail concerning the technical and institutional processes involved in electricity supply in the UK. The purpose of supplying this detail is to provide an accurate sketch of the technical challenge involved in delivering an industrialised society’s energy requirements. This is sketch will help in later chapters to delineate the boundary between the technical and—

as it will be argued—far more substantial social barriers to certain directions in sociotechnical development, especially where sustainable alternatives to current patterns are concerned. Finally, Chapter 2 examines the regulatory and legislative environment in which electricity co-provision operates commercially. Notwithstanding the criticism, above, of simplistic discourses on technology change—which emphasise market- delivered innovation to the exclusion of all else—this is still an essential component of a broader discussion about processes of sociotechnical change.

Chapter 3, Sociotechnical Transitions, applies a transitions theoretical framework to this empirical analysis of co-provision technology in the context of existing technical and

(16)

6

institutional arrangements for electricity supply in the UK. It brings in specific energy co-provision projects as examples of how the theoretical concepts provided translate into practice. Among these is a project which is experimenting by networking the electricity supplies of around one hundred homes of council house residents in an economically deprived area, in order to allow them to function as a microgrid. Around 60 of these houses will have rooftop solar panels installed. This microgrid allows energy to be moved around networked houses in order to utilise a higher proportion of the total electricity generated by the solar panels. This saves residents money, but also takes pressure off the national grid by substantially reducing the peak demand it has to supply to the networked houses. This practice, called ‘customer group aggregation’ is used in a similar way in another co-provision project examined in Chapter 3, which allows integrated homes to become a larger consumer. This larger consumer can take advantage of remuneration offered by the national network operator for shifting load away from times of peak demand. Chapter 3 systemises information about these and, other, energy co-provision projects. This information is put into a framework informed by the theoretical concepts supplies by tenets of sociotechnical transitions theory.

Chapter 4, Technology Choice, discusses the implications of the empirical findings set out and analysed in Chapters 2 and 3. The chapter’s major theme is the governance of sociotechnical change. Crucial to the discussion is the notion that technological innovation—i.e. the invention or repurposing of technology as a means to reorder technical systems—cannot be considered a politically neutral act in the light of the inherently social nature of technology and technological change (Hilgartner 2009). It argues that the dominant discourse in technology governance at present employs a unitary conception of ‘progress’ which acts to occlude the possibility of alternative sociotechnical paths (Stirling 2015), and that this leads to there being a democratic deficit in processes governing longitudinal sociotechnical change. In doing so the discussion broadens to consider what the notion of multiple possible innovation trajectories, and the research and policy discourse which shapes them, means for sustainable development more generally. It attempts to couch the concepts constituent to this discussion in terms used in a sustainable development framework. The chapter concludes by locating the discussion of technology choice and the research and policy discourse in processes governing sociotechnical change, along with the findings

(17)

7

presented in previous chapters, within ongoing research in the field of science, technology and sustainability.

The major aims of this thesis are, in short, to challenge simplistic discourses on technology change, and to frame the topic as an issue of societal transition of upmost importance to environment and development research. Within this, the major themes are climate change and environmental degradation caused by a societal commitment to fossil fuels, poverty and inequality linked to—or compounded by—current arrangements for the supply of electricity, the inherently social nature of technology, and issues of participation and democracy in the processes that govern sociotechnical change. These are all broad themes, none of which could be explored thoroughly on their own in a thesis of this length and scope. More concretely, then, the thesis sets out to conclude by answering the following research question and sub-questions, based specifically on the UK experience of energy provision and sociotechnical change:

Research Question: What dynamics underlie processes of sociotechnical change affecting systems of energy provision, and what theoretical and analytical tools are there for the study of these dynamics?

i. how can these analytical and theoretical tools be applied empirically in the case of co-provision technologies, and what insights does this offer?

ii. do these tools have a normative application for designing policy which will aid the diffusion of co-provision technologies?

iii. what insights can theories of sociotechnical change offer to broader inquiry into societal change, especially with regards to sustainable development?

The remainder of this chapter starts by outlining the methodological approach taken for this study. Following this, it sets out some historical background to the study. This provides some context both of electricity systems and their technical and institutional development, but also for the theoretical approaches to studying them.

1.2 Methodological approach

This study attempts a qualitative analysis of a dangerously quantified topic. Dangerous in the sense that, due to the enormous wealth of quantitative data on electricity systems, it is surely possible to contradict all of the factual claims—those based on numerical

(18)

8

information—made in this thesis. Interpreting the mass of data on consumption (MWh), installed capacity (watts), retail cost per unit (£/KWh), spinning reserve ramp-up time (seconds), transmission efficiency (%), load and capacity factors (%), grid frequency (Hz)—along with various other energy system-related metrics associated with the topic of renewable energy and co-provision—could well fill many theses. Indeed, earlier drafts of this thesis contained far more technical and numerical information than does this final version; the major themes were becoming lost in pages of technical description and supporting numerical information that has now, although not gone entirely, been vigorously edited.

The paring down of quantitative data was not because of a conviction that these numbers do not matter, they most certainly do: they provide important indicators of the performance of a system upon which all of society’s wellbeing relies; understanding and responding to them is a key part of a decision making process that affects—and should include—everybody. And herein lies the problem with over-emphasising quantitative data such as performance indicators on electricity systems. To aggregate these data and present them as ‘stark facts,’ natural and undeniable, is to take them out of the social context—the complex “network” of relations (Berteaux 1981)—that gives these figures significance (Bruner 1990). The danger—to evoke the term a second time—of doing so is that the process of decision making around society-shaping technical systems becomes subject to the hegemony of aggregate numerical models.³ Technical decision making processes, which should be understood for what they are, broad sociotechnical dialectics, are reduced to a matter of technical realpolitik.

While, then, an observation might be made in quantitative terms about, for example, the comparative flexibility of nuclear and coal generation, or figures on the average capacity factor of wind generation in the UK may be cited in order to support a point, the conclusions of this thesis are not based on in-depth analysis and modelling of electricity system performance metrics. Nor does this thesis speak from a position of expertise on such matters. Indeed, quite the opposite is true. The researcher has very little technical knowledge of systems of electricity provision further to that which was gained as part of the process of conducting this study. The key method behind the

3 This is a point examined further in Chapter 4, which examines the effect on processes of sociotechnical change of the tendency in modern politics, following the study of economics, to reduce technological decision-making to consumption choices by aggregate choice-makers (Miller 1995).

(19)

9

deployment of quantitative technical data has been to exercise caution in weighing up different sources’ collection and presentation of data, and to cross-check to ensure that this thesis’ reproduction of these data is intellectually honest in its ascribing of significance to it.

The qualitative research findings presented in this thesis—observations of the sociotechnical phenomena analysed and the theoretical tools acquired to do so—were collected in a study comprised of two components. One of these was a field study comprised of a series of interviews with expert informants. The second research component of this study was an extensive and far-reaching desk study of academic literature, policy documents, patents, and some related articles in popular journals. This second study served both as triangulation for the data gathered in the field study, as well as providing information on topics not covered directly by interviews with expert informants. Both of these two components are briefly outlined and reflected upon below.

1.2.1 Field study

The field study was conducted over a few months between July and early September 2015, but largely consisted of two stages. The first stage, in July 2015, consisted of one informal conversation, which provided the leads required to find—and make initial contact with—more informants. There was an intervening period in which there was some email correspondence between the researcher and all of the informants. The second substantive stage was in a two week period, between late August and early September, in which all four formal interviews were conducted. Each lasted between 35 minutes and one hour. All four informants were associates of the initial contact, who was an associate of somebody known to the researcher. With their consent, this contact gave the researcher email contact details for the four informants. Introductions were left for the researcher to make. Of the six contacts invited with bespoke email introductions—outlining specifically why they had been selected to invite to the study—four responded positively and an was interview arranged, one expressed interest but was unable to find time to participate, and one did not respond.

The purpose of the field study was simply to discover what insights into electricity co- provision could be gained by talking to expert practitioners in the field, all of whom

(20)

10

were associated with one of the co-provision projects examined in Chapter 3.

Accordingly, there was no interview guide as such (but see Appendix for list of broad question categories), and the interviews took the form of informal conversations in which the informants were invited to talk about their academic backgrounds and professional histories, and to explain their current work surrounding the implementation of electricity co-provision. Co-provision is an analytical term found in academic literature. None of the informants volunteered this term in relation to their work in the field of energy provision more widely. All did, however, give detailed accounts of their work that do match up very closely and un-problematically with definitions of electricity co-provision found in the literature (see for example Watson 2004). One of the initial major insights gained from the field study, then, was affirmation of the empirical validity of the theoretical notion employed by this thesis, originally gleaned from the academic literature.⁴ Because of the function these interviews performed as

‘expert aids,’ as opposed to data for interpretation, they were not transcribed in full, but recordings and notes were kept for the duration of the writing-up process.

On reflection, these interviews may have been conducted too early in the research process. At the time of the interviews the researcher had failed to grasp the fine details of the technical and institutional processes involved in electricity provision in the UK, or indeed of the broader sociotechnical concepts underlying co-provision. This understanding has been pieced together since the time of the interviews using the desk study (see below) and the interview notes themselves. While this could be presented as an extra insight gained as a result of the field study, the researcher’s lack of a detailed understanding in this area—which could have just as easily been rectified by a desk study and without field interviews—acted, during the interviews, as a barrier between researcher and informant. The informants were all busy professionals taking time out of their day to aid what they had been led to believe was research that would support environmental causes⁵; they all came to the interview with overly-optimistic expectations—as it turned out—of the researcher’s knowledge of the broader electricity

4 This was not the notion the researcher had adopted in order to understand and explain co-provision technologies at the time of the interviews, however. This insight was pieced together much later while going over interview notes and relevant literature.

5 The email inviting informants to participate in the study explained that the purpose research was in the area of environment and development.

(21)

11

industry, which led to valuable time being taken up with their having to explain what are—in light of the researcher’s present understanding—basic concepts.

All of the informants were very patient in this regard, but the result is that opportunities for the researcher to elicit deeper insights into co-provision from the informants were, in all likelihood, missed. The informants, as the parties with far superior knowledge of the area, very much had to lead the discussion. This was of course to be expected and—to an extent—welcomed, but the informants’ expressions and body language would sometimes register disappointment when the researcher failed to react to the significance of something they had said. Hindsight now shows that some of these statements were indeed quite extraordinary in content, and warranted a reaction of sorts.

On other occasions the recordings show the researcher reacting to what is, in fact, a rather mundane detail. This would have no doubt been perplexing to the informants.

Allowing the informants to lead the discussion has, on the other hand, had many benefits as well. They had the freedom to talk at length, which all of them did to varying degrees, and a vast number of insights contained within the recordings of these interviews has emerged steadily over the course of their use as writing aids: every time the researcher returns to these recordings having read further into the topic’s web of complexities, these recordings yield a fresh, sharper understanding that draws together previously-scattered dots.

It is impossible to say definitively, but had the researcher been more prepared for these interviews, better able and more confident to guide the discussion, the course they may have taken could have been very different and offered a different set of insights. In this regard, it would perhaps be untrue to state that on reflection the researcher would not have conducted interviews at this stage of the process: the field study has proved invaluable to this research project as a whole, despite these caveats. Ideally, however, there would have been a second round of interviews with all of the informants at a later stage in the process after the extensive desk study was completed. This would have been especially interesting in this case, given the regulatory developments that have occurred between the time of the interviews and the completion of this thesis (see Chapter 2).

Time constraints on the researcher have certainly been one factor in why this second

(22)

12

round of interviews has not taken place, as has the practical difficulty of scheduling interviews with busy informants to fall within a short timeframe in the UK.⁶

Upon further reflection, however, what has also been a barrier to securing a second round of interviews has been the researcher’s sense of embarrassment at the lack of knowledge demonstrated during the interviews. This embarrassment may seem illogical, given that the interviews were explicitly organised with ‘expert’ informants who would undoubtedly have superior knowledge and understanding of the topic. Further deconstruction of this sense of embarrassment reveals that it emanates from the researcher’s approach to handling this knowledge gradient, rather than the fact of this gradient itself. As opposed to being upfront with the informants about the researcher’s level of technical and empirical knowledge of the UK energy industry, and the highly social nature of this research, the researcher tried to ‘wing it’ by showing few signs of not understanding certain aspects of what the informants were saying, and confidently asking questions.⁷ This decision was, again, taken out of a sense that to admit to such a poor grasp of the area would cause the informants’ consternation at having been asked to take the time to explain basic elements of their work. This concern was most certainly misplaced, and all of the informants were amicable, understanding individuals;

several of them recalled during the interview conducting similar research for their master’s research project, and were just happy to help in whatever way they could.

Overall then, the field study provided a valuable insight to the topic, has to a significant degree structured the way that this study later developed, and has been deeply formative with regards to this thesis. The major ethical considerations for conducting the research—namely the informants’ valuable time taken by the interviews—although real, perhaps played more on the mind of the researcher than they need have done, in the sense that they affected researcher’s presentation in interview in a manner that proved to be counter-productive. A lesson for future research, then, is to find a better approach to handling knowledge gradients between researcher and informants so that un-wise and misleading strategies do not become a barrier—in the researcher’s mind—to further contact. An important element in successfully embodying this lesson will be greater confidence. A researcher must both have confidence that the research they are

6This in the interests of only making one trip from Norway.

7 Some of these questions, viewed in the light of new understanding reached since the interviews, are somewhat inane.

(23)

13

conducting is valuable, and acknowledge that a knowledge gradient when speaking to experts in another field is inevitable. Gaining this confidence is a matter of reflecting on this realisation, and ensuring that preparation for interviews is structured and thorough.

For a researcher particularly interested in the sociotechnical phenomena, this is undoubtedly going to be a salient issue in future; and in this sense it is in itself a valuable insight to have come out of this study.

1.2.2 Desk study

The desk component of this study has been going on around and throughout the process of planning the field study, writing up field notes, analysing these notes, and writing up this thesis. Indeed, it continues now, and will do for some time to come. It has been far- reaching and encompassed a broad range of literature and documents. For the purposes of a brief explanation, this broad range of literature and documents can be divided into four main categories—all still exceedingly broad themselves—all of which are outlined briefly below. The first of these categories could be described as academic literature that falls under the loosely-defined heading of environmental policy literature. The second, more clearly defined category is the sociotechnical transitions literature, used at length to develop this thesis and relied upon heavily for the analysis in Chapters 3 and 4. The third category is a somewhat miscellaneous one, which throughout the writing process of this thesis has been comprised of the documents organised into a desktop folder simply called “Reports.” It contains articles from popular journals, government communications, reports written by professional associations, think tanks and industry actors, along with all other documents that do not fit into any other of the four categories. The fourth category contains academic literature and other documents—such as patent documents—that pertain to technical and engineering topics.

Environmental policy literature is a diffuse category. It is comprised of a wide range of research papers and other academic literature that has a bearing on environmental policy, some of these bearings being more oblique than others. Key literature in this category includes longer works such as The Provision of Public Services in Europe, edited by Hellmut Wollmann and Gérard Marcou (2011) and Reclaiming Public Ownership by Andrew Cumbers (2012). Both of these books blend perspectives on political economy and the study of policy to deliver an empirically robust theoretical

(24)

14

framework for contemplating the nature of public services and utilities as sociopolitical institutions. A series of papers produced by the Science Policy Research Unit (SPRU)—

some of which also deal with sociotechnical transitions theory (e.g. Berkhout et al 2003, see below)—and the Social, Technological and Environmental Pathways to Sustainability (STEPS) Centre also formed a key part of the environmental policy literature examined. Chief among these were Stirling (2009) and Stirling (2015).

Alongside these, papers such as Elizabeth Shove’s (2009) Beyond the ABC: climate change policy and theories of social change offer a critique of policy interventions that are blind to the coevolution of technical systems and society. Finally, for this category, were papers that examine the developing institutional forms of distribution in energy, and which provide this thesis with its notion of co-provision. Chief among these was Watson (2004).

The books and papers that informed the chapter in this thesis on transitions theory do not comprise a complete literature review of the transitions literature, although they are able to offer a semblance of the contours within this area of thought. Thomas Hughes’

(1983) Networks of Power: Electrification of Western Society 1880-1930, although not technically part of the transitions literature, was formative for the understandings underlying much of this body of work. Other works by historians of technology, such as David Nye’s (1990) Electrifying America: Social Meanings of a New Technology also fall into this category. Another branch of the coevolutionary approach—that, for a while, developed alongside transitions theory, before merging with it (see Geels 2004)—and which also built on Hughes’ approach, is the study of large technical systems (LTS). The Governance of Large Technical Systems, edited by Olivier Coutard (1999), includes some chapters that have been very influential in this thesis. Chief among these are Coutard (1999), Summerton (1999), and Guy et al (1999).

Of the transitions literature itself, a very influential early contribution as far as this thesis is concerned is Rip and Kemp (1998), which draws together from diffuse sources many of the concepts that now underpin a holistic conception of transitions theory.

After Rip and Kemp, the literature branches out. Again, in terms of contributions which have proved influential in this thesis, Rotmans et al (2001); Geels (2004); and Schot and Geels (2008) refine the coevolutionary multi-level perspective understanding of transition (see Chapter 3). A conscious and productive effort to incorporate civil society

(25)

15

action into transitions approaches is found in Seyfang and Smith (2007); Seyfang and Haxeltine (2010); and Hielscher et al (2011). Meanwhile, Berkhout et al (2003) and Smith et al (2005) address some issues of ontological perspective which arise as a result of the way transition theory develops and is applied to empirical examples. Geels et al (2015) provides a comprehensive inductive analysis of the UK electricity industry using a branch of transitions theory and, finally, Johnstone and Stirling (2015) make interesting adaptations to the approach in light of research findings of an in-depth study comparing the network of actors and processes informing the civil nuclear industries in the UK and Germany.

The third category of literature involved in the desk study contains documents too numerous to list individually here.⁸ The methods used to accrue this large number of files and papers began as rather ‘scatter-gun’ web searches, but became more refined over time as the researcher became both more in-tune with relevant key words, and more accustomed to how publications on government departmental websites are organised and categorised. The fourth and final category of documents was perhaps the most difficult to approach and engage with from a social research perspective. Walker (1986), and Sioshansi and Davis (1989) were a useful way into productively linking technical information with the social dimensions of energy supply, as this is also the approach both these papers take. More straightforward technical articles on microgrid architecture and controls were Lasseter et al (2002); Short et al (2007); Zamora and Srivastava (2010). Markvart (2006) provides a brief and accessible introduction to the topic.

The literature listed here does not constitute a holistic literature review on any particular topic or family of theory. The presentation here of the influences from literature on this study is intended, rather, to demonstrate the methodological—if not methodical—

approach taken to literature. Selecting and dividing documents into the four broad categories outlined above was in fact an ongoing, iterative process that helped to inform the writing process. The weighting of source type and topic in the list above reflects very much the weighting these sources are given in the thesis as a whole. This thesis strives for multidisciplinarity, but in doing so risks doing injustice to all of the disciplines it calls upon. As such, the same disclaimer as was used in relation to the

8 They are, however, referenced throughout this thesis, and can be found in the reference list.

(26)

16

deployment in this thesis of technical data applies to the use of theoretical frames and concepts. They are not intended to provide a comprehensive overview, although they selected judiciously and with care taken to not oversimplify or misrepresent any concept or theoretical approach in the process.

1.3 Background: electricity supply as the technical embodiment of social, political and economic circumstances

Societal requirements of energy, water and sewage, telecommunications, and transport are provided by exceedingly complex organisations of technical artefacts, institutions, social relations, and large amounts of capital operating in network (Coutard 1999).

These organisations are crucial to maintaining social and economic life in industrial societies. Because of these systems’ technical nature and capital intensity, it is easy to designate them an issue of simple technical-economy, where engineering quality and pricing structures are the only factors to have a bearing on the daily lives of societies served by these systems. Starting with Hughes (1983), however, the broader links between technical systems and society have become a subject of study by the social sciences. Underlying this new interest, Olivier Coutard suggests, is the assumption that

“the relations between these socio-technical systems and ‘society at large’ are essential to understand the evolution—the coevolution, one might say—of both the systems and the society” (Coutard 1999, 1, original emphasis). It is this process of coevolution with regards to electricity provision to which the below now turns.

1.3.1 Electrification: a process of system-society coevolution

Coevolution, then, in a sociotechnical systems sense, means that social, political and economic circumstances shape electricity supply and become embedded in the physical hardware and infrastructure attendant to electricity supply systems; characteristics which themselves have a profound bearing on the arrangement of social and economic life. Understandings of how society shapes technical systems of utility provision are missing from analyses that assume sociotechnical processes—such as electrification in public spaces, businesses, and homes—are extra-societal phenomena which proceed in inevitable fashion along ‘natural’ lines. This conception of sociotechnical innovations,

(27)

17

such as electrification, as a ‘natural’ process—the ‘implacable force’ understanding—is at odds with the understanding of electrification as “a social process that varies from one time period to another and from one culture to another” (Nye 1990, ix). This is a crucial distinction for a great number of reasons, but is particularly salient for environment and development research concerned with energy provision, as policy discourses that are couched in an ‘implacable force’ understanding of technological change can in themselves constrain possibilities for ambitious and meaningful moves towards greater sustainability (see Chapter 4).

David Nye’s study of the early electrification process, as it occurred in a small town in the United States, is illustrative of how social, political, and economic circumstances can structure and become embodied in the technical arrangements of electricity supply.

Nye presents how “in the years between 1890 and 1920 a dialogue on electricity took place in American society,” where “some voices were louder than others” (1990, 138).

This dialogue took place as Americans tried to decide

“whether to construct many small generating stations or a centralized system, whether to use alternating current or direct current, whether to place ownership in public or private hands, whether to establish rates that favoured the small or the large consumer, whether to give control over the system to technicians or to capitalists or to politicians” (ibid).

The experience in Muncie, Indiana, described by Nye is analogous with experiences of electrification elsewhere in America, Europe, and further afield; but as Nye explores, local societal contexts varied hugely leading to diverse sociotechnical outcomes of this dialogue. There was during this period, for example, a notable difference between how electrification proceeded in the US and in parts of Europe. In the former, domestic electricity surfaced unevenly and slowly, and was viewed as a lavish and conspicuous means with which the wealthy elite could define itself; in Scandinavia, Germany, and Holland, on the other hand, a political commitment to universal domestic electricity provision led to there being only one in ten homes unsupplied by 1930 (ibid, 140). The electric lighting of public spaces, in contrast, showed an almost inverse pattern: in the US spectacular lighting arrays in streets and shop windows by advertisers and retailers were a part of rapid growth in the electric lighting of public spaces—led by private enterprise—which “quickly developed far beyond social necessity” as electrically- powered spectacle became constituent in corporate-public relations (ibid).

(28)

18

These contrasts reflect the differences in the political-economic and social contexts of electrification of the US and Europe. Nye suggests that the decentralised nature of governance in the US meant that top-down policy goals, such as the universal provision of domestic electricity, would have been difficult for central government to implement across the many jurisdictional boundaries of a federal nation (ibid). This difficulty was compounded by the lack of appetite for big government welfare interventions and a capitalistic—in comparison to Europe—ideology: at the time the railways and telephone networks in the US—which were publicly owned throughout Europe—were privately owned (ibid). The commodification of electricity in the US during this time reflected other aspects of American society too. Although there were competing notions of what electricity was, and how it might be integrated into society, much of the early twentieth century discourse around it suggests a “popular understanding of electricity as a quasi- magical force with utopian potential” (ibid, 157). It is, perhaps, understandable that state control of this technological miracle, no less—eulogised in the form of public spectacle—was anathema to a large number of Americans, given the particularly strong currents of laissez-faire individualism in early twentieth century American society.

The distribution of ownership and comparative rates of domestic electrification in the US and Europe are just one example—painted here in broad strokes—of how different social, political and economic circumstances can shape technical systems. The point to emphasise is that “in no society was electrification a ‘natural’ or a ‘neutral’ process;

everywhere was it shaped by complex social, political, technical and ideological interactions” (ibid, 139). The choices made—in this case surrounding ownership and control of the developing technical system—structured, to a large degree for many decades to come, the development of that system. As stated, these were not the only choices confronting societies in the process of deciding how to integrate electricity generation-transmission technologies into social and economic life, and nor were these choices made on the basis of equal representation of competing views. The political struggles that led to sociotechnical decisions which influenced subsequent developments were characterised by differences in power between interests. The process of politicised sociotechnical decision making that led to the emergence of centralised bulk power as the form of electricity provision in the UK is explored below.

(29)

19

1.3.2 Early system expansion

System expansion can be theorized in a number of ways. Perhaps most straightforward is the observation that technical systems must sometimes expand spatially in order to increase the level of coverage. The public service obligation attached to electricity provision in some cases—such as parts of Europe in the early twentieth century—is one factor in why supply systems must increase coverage through expansion; alongside this factor is the normal economy of scale factor which applies to many capital-intensive activities such as electricity provision. Additionally, however, electricity supply systems also expand across functional—as opposed to spatial—boundaries in order to capitalise on more varied load types; this improves the load factor and increases the efficiency of supply (Summerton 1999). This drove much of the early system expansion in the US, for example, where electric traction companies in the late nineteenth centuries became some of the largest suppliers of electricity to municipalities (Nye 1990). However, as Summerton notes, some theoretical positions hold that “sociotechnical networks are to be understood as Machiavellian instruments of power” (Summerton 1999, 95). From this perspective, system expansion could be viewed not only as a necessary public service or an exercise in technical-economic expediency, but also as a political tool for domination. There is certainly some evidence of attempts a domination by the interests party to the political struggle in the US and Europe known as ‘the battle of the currents;’

of which Hughes (1983) provides an insightful account.

Borrowing from military terminology, Hughes describes the development of electricity supply systems in terms of an advancing line—or front—that is frequently beset by

‘reverse salients:’ sections of the line or front that fall behind the rest (Hughes 1983).

Once a reverse salient in a developing technical system is identified it can be defined in the terms of a ‘critical problem’ (ibid). A critical problem is a specific design flaw, missing technological attribute, or knowledge gap that is causing the system’s reverse salient. Solving a critical problem requires a new invention or new knowledge that can bring a lagging-behind section of the system back into line (ibid, 80). Typically, to describe a system component as ‘lagging-behind’ is to mean that it works economically at one scale but is still inefficient enough to be uneconomical when operating at a larger scale; this is why a reverse salient is often revealed as a system expands or, to extend the metaphor, begins its advance (ibid).

(30)

20

The so-called ‘battle of the currents’ arose from the spatial expansion of an electricity supply system hampered by a reverse salient. As generation technology became more efficient and reliable, and as demand for electricity grew, managers operating electrical supply systems identical to that patented by Thomas Edison attempted to connect to their distribution network users spread over a larger geographical area (Hughes 1983).

Edison’s system was able to deliver electric lighting to a compact urban area for a cheaper per-unit price than natural gas. When expanded spatially in this manner, however, the transmission losses became so great that the system was no longer economically viable (ibid). This is an example of a reverse salient because the transmission of electricity was just one system component of several, but the only one failing to keep pace with the technological and spatial development of the system as a whole. The critical problem defined in this instance was that the direct current (DC) generators used by Edison’s original system could not be viably transformed between high voltages for efficient transmission and low voltages for safe distribution; solving this problem required the invention of new transformers and transmission hardware (ibid).

Hughes asserts that this process—system expansion, identifying a reverse salient, defining it as a critical problem, and inventing the solution—results in “innumerable (probably most) inventions and technological developments” (ibid). As opposed to inventing from scratch new technologies to fulfil an as-yet unknown or at least not widely-practiced purpose, inventors spend ‘most’ of their time solving critical problems with inventions that will realign reverse salients in a system. In doing so inventors will sometimes invent the kernel of an entirely new system (ibid, 91). This was the case with the DC transmission issue: in trying to find a way of making long-distance transmission viable inventors developed the alternating current (AC) transformer and its attendant transmission components.

For some years after the invention of the AC system there was sustained competition between AC and DC system manufacturers for market share and technical acclaim in academic engineering circles (ibid). Hughes cautions against assuming that, just because polyphase AC transmission-distribution is the worldwide standard today, its inherent technical-economic advantage over DC means its eventual victory in the so- called ‘battle of the currents’—which ensued from its invention—was always inevitable

(31)

21

(1983, 81). Crucially, this inherent technical-economic advantage of an AC system was not and–-as is discussed in Chapter 2-–is not an assured matter of universal fact but of sociopolitical, legislative and economic circumstance (ibid). Because the comparative technical-economic advantages of both systems were open to debate, competition was intense. Interested parties resorted to using “political power” in an attempt to settle the contest (ibid, 107). For example, in an attempt to gain commercial advantage over the insurgent alternating current, US-based manufacturers of direct current systems went as far in 1889 as successfully lobbying the State of New York to replace hanging with the AC-powered electric chair in executions. The hope was that this gruesome association would harm alternating current system sales (ibid).⁹

1.3.3 The shift to ‘bulk power’

While the political power struggles wore on, alternating current transmission- distribution systems continued to steal market share from direct current systems, which could not compete when supplying electricity over a larger geographical area. This meant that while alternating current systems were a clear 'winner' in rural areas, in urban areas debate as to whether or not direct current might be more suitable continued into the twentieth century (ibid, 81). In the UK the urban-rural distinction which emerged between systems was intertwined with another dichotomy: between small scale, technologically heterogeneous systems and large, standardised systems centrally supplying a large area (ibid). Although among engineers, especially in the US, support for the latter was almost ubiquitous by the beginning of the twentieth century, in London small scale DC generation-distribution networks continued to prevail.

Again here, this debate about the most appropriate scale for the deployment of electricity generation-distribution networks was not purely technical, but also political.

The fact that “these small scale undertakings were appropriate for the demand they

9Not only this, but one Harold Brown, an employee of DC-advocating parties, attempted to demonstrate to the general public the existential threat posed to them by AC using another macabre method. On more than one occasion Brown—in front of an audience—tortured dogs with non-lethal doses of direct current before killing them with an alternating current. He was not, however, an engineer but a spin doctor, spearheading DC-interested parties’ “nontechnical compensatory response” to the commercial threat posed by a rival system (Hughes 1983, 107). The ‘demonstrations’ were part of a wider lobbying campaign—which failed—designed to bring about legislation that would limit the voltage of AC transmission to 300 volts, thus removing its technical-economic advantage over long distances (ibid, 108).

(32)

22

met¹⁰ and for the legislative circumstances of the day” (ibid, 234) instilled confidence in the “managerially conservative interests” that sought to protect the autonomy of local authorities and contributed to the preservation of the small-scale style of electricity supply in London in the first decade of the twentieth century (ibid, 255). The power wielded collectively by these interests was considerable. So much so that the multiple efforts by celebrated up-and-coming engineer Charles Merz to persuade parliament to allow him to interconnect and standardise electricity supply across London were all confounded. After one such occasion in 1905, David Lloyd-George, MP and President of the Board of Trade, reportedly had to console the somewhat perplexed industrialist:

“my dear young friend, this is not a question of engineering, it is a question of politics”

(Hughes 1983, 251).

Ultimately, it was the unprecedented demands placed on the electricity supply system by World War I that caused the political opposition to legal change to give way to system interconnection, standardisation and 'bulk power'; this post-war trend culminating in the establishment of a national grid in 1926 (ibid). As the remainder of the small-scale, local infrastructure for electricity generation-transmission was replaced with a large-scale national standard the ‘battle of the currents’ too came to a conclusion with the standardisation of alternating current two years later in 1928. The observation that “history is written by the victors” could well apply to this commercial struggle:

Hughes points out that the fact direct current continued to operate commercially in some areas right up until it was swept away by standardization is often overlooked by historians (1983, 85). Instead, the modern narrative of the ‘battle of the currents’ holds that direct current generation-transmission was out-performed by its rival and died an inevitable death brought on by technical-economic inferiority.

Of course, this is not to say that without standardisation-through-regulation direct current would have survived as a commercially viable option in the context of twentieth-century power needs. As noted, World War I was a catalyst for sociopolitical change which allowed for a shift in industrial logic towards ‘bulk power’ that required the long-distance transmission of centralized generation and more heavy manufacturing capacity, both of which required ‘stronger’ AC power. Rather, the point is that while direct current generation-transmission ceased to embody the sociotechnical ideal in the

10 Namely, these were dense, urban areas.

Society, Technology, and the Sustainable Provision of Electricity